#9 Pi biased by read coverage?

[Anonymous]

Originally created by: kcde... (code.google.com)@gmail.com

Hello,

Is it possible to remove genomic regions from analysis when using Variance-sliding.pl based upon read coverage within the window? I am worried that Variance-sliding.pl calculations show low values of Pi, Theta, or D, when in fact there is just low read coverage in the specified region. Is there a way to correct these statistics with percentage of the window that is covered by reads, or remove these regions from the analysis altogether?

Any discussion of how read coverage within a window will effect these statistics, or can inform window length, would be greatly appreciated.

Many thanks,

Kevin

[Anonymous]

Originally posted by: RoKof... (code.google.com)@gmail.com

Hi Kevin,
you can subsample to an uniform coverage, there is the script subsample coverage. This also removes regions with a coverage lower than the targeted coverage
cheers ro